Eur J Forest Res (2014) 133:611–622 DOI 10.1007/s10342-014-0789-3

ORIGINAL PAPER

Economic assessment of use values of near-natural compared with rotational forestry in

Dareskedar Workie Amsalu • Jette Bredahl Jacobsen • Thomas Hedemark Lundhede

Received: 21 June 2013 / Revised: 12 January 2014 / Accepted: 24 January 2014 / Published online: 2 February 2014 Ó Springer-Verlag Berlin Heidelberg 2014

Abstract The present study is a cost-benefit analysis of Introduction converting the current rotational forestry (RF) of Norway spruce stand into near-natural forestry (NNF) of , Conversion of the traditional rotational forestry (RF) of Nor- based on two representative soil conditions and visitors way spruce (Picea abies) into a more ecologically oriented popular case areas in Denmark, considering welfare eco- and society-centred multi-functionality forestry approach nomic values of timber, recreation provision, and ground- called near-natural forestry (NNF), using native broadleaved water recharge. The study answers the major research species and mainly European beech (Fagus sylvatica), has question of how large the welfare economic values of been growing across many European countries. There are two recreation and groundwater benefits of the conversion are underlying common factors for the conversion (Spiecker et al. as compared with timber benefits. The net present values 2004). First, with rising living standards over time, modern (NPV) of the benefits were calculated for an infinite time society’s interest in multiple functions of forest, such as out- horizon at a 3 % discount rate. The results reveal that door recreation provision, water protection, converting into NNF would result in a NPV of at least conservation, and carbon storage, increases, and NNF is often 6,832 € ha-1 from use values of recreation and water better at securing these ecosystem services than RF. Second, benefit on a site with good soil and a high visitor fre- the traditional monoculture management of Norway spruce quency, as is typical in the eastern part of Denmark. On a has faced severe damages by storm, insect pests, and other site in the west of the country, with poor growth conditions disturbance. and a lower visitor frequency, the gain is still substantial, Despite lively discussion of the development of society- namely 5,581 € ha-1. These benefits though come at a cost centred and nature-based forestry policy, the welfare eco- of 3,375 and 6,206 € ha-1 from timber production, nomic impacts assessment of the expected social and respectively. This means that the economic value of use environmental benefits of the policy is yet at the young values of recreation and water benefits outweighs the loss phase. The current study, therefore, aims at providing an of timber on good soil conditions but not on poor soil economic assessment of the socio-economic effects of conditions. converting RF of Norway spruce stand into NNF of beech, based on two representative cases at each end of the Keywords Cost-benefit analysis Near-natural forestry spectrum of soil conditions and visitor frequencies in Recreation score Tree growth Run-off Water Denmark. We consider welfare economic values of the Conversion most demanded forest functions, namely timber production and two non-timber benefits (recreation provision, and groundwater recharge and protection). Conversion from one forest management into another, including a tree spe- Communicated by M. Moog. cies change, is a long-term change, and many benefits may first be seen much later. Therefore, an important attribute D. W. Amsalu (&) J. B. Jacobsen T. H. Lundhede University of Copenhagen, Copenhagen, Denmark of the study is that we make the flow of the non-timber e-mail: [email protected] benefits age dependent. 123 612 Eur J Forest Res (2014) 133:611–622

Apart from use values, forests also provide non-use Table 1 The number of visitors and soil conditions in two study values. However, the aim with the current study is alone to forest areas in Denmark estimate the use values which are more tangible. Site Gribskov Klosterheden Source The study would give an important insight for policy characteristics makers and forest managers to scheme site-specific eco- Number of 78 (448,000 visits 25 (154,250 visits Jensen nomic instruments, such as payment for ecosystem services visitors over 5,698 ha) over 6,265 ha) (2003) (PES), so as to attain sustainable forest management. (visits The existing forest economics literature that has ana- ha-1 year-1) lysed the business economics of NNF compared with RF Annual 613 850 DMI finds ambiguous results. Tarp et al. (2000) and Nord-Lar- precipitation 2012 (mm) sen et al. (2003) find NNF to be superior, mainly due to Location Northern Northwest Jutland reduced establishment costs. Nord-Larsen et al. also attri- Soil condition Good soil class Poor soil class Møller bute the result to more optimal harvest timing. Other For beech site class For beech site (1933) studies find the profitability of conversion into NNF to be 1(H50)of24m class 4 (H50)of inferior to RF in some stable conditions (Jacobsen et al. (top height at age 17 m (top height 2004; Duncker et al. 2011). Yet, the business economics of 50). at age 50) the conversion is only one side of the coin. The other is the For Norway spruce For Norway spruce non-marketed ecosystem services. site class 2 (H50) site class 3 (H50) Price and Price (2008) throw light on the economic cost- of 25 m of 22.5 m benefit analysis of continuous cover forestry or NNF of conifers trees with a monetary estimate of recreation benefit and water evapotranspiration loss against the RF system, with respect to soil conditions and visitor frequency assuming the benefits are constant over time. However, the (Table 1). The conversion is meant to achieve two major recreational values are known to depend on age and forest objectives: wood production and recreation values on both structure (Boman et al. 2010). Consequently, we examined sites (Larsen and Nielsen 2007). An important co-benefit is the dynamic welfare economic values of recreation benefit of an increased amount of groundwater produced as broad- different phases over a rotation of NNF beech. This was leaved trees evaporates less. The forest areas on both sites possible by tracing Danish recreationalists’ willingness to are composed of different tree species, mainly Norway pay (WTP) for attributes of NNF of beech (Nielsen et al. spruce, European beech, different species of (Quer- 2007) to different phases of the stand development, using the cus), and Douglas fir (Pseuudotsuga menziesii) (Rune respective ‘‘recreational scores’’ developed in the EFOR- 2009). WOOD research project (Edwards et al. 2010). To make sure that we compare initiatives originating Groundwater is an important resource in Denmark, from the same state, the starting point for the developed fulfilling almost the entire water demand. In densely pop- scenarios is assumed to be bare land after the existing ulated areas, the resource is becoming scarce, and matured Norway spruce stand is clear-felled. Due to risk of increased demand for water is therefore an issue. The windthrow in Denmark, it is normally not considered fea- provision of groundwater underneath forests depends on sible to regenerate under shelter, and natural regeneration age as does recreation (Boman et al. 2010). However, the is not possible due to lack of seed trees in the conversion biggest difference lies in the species change and thereby phase. The ‘‘baseline scenario’’ is a continuation of RF of the amount of water running through. Norway spruce using clear-felling and replanting in a The remainders of the paper are organised as follows. 60-year rotation period on the good soil and 70 years on Section two discusses definition of the traditional and the poor soil, for an infinite time horizon (Forest Eco- conversion scenarios, methodology and economic calcu- nomics Table 2006). At the time of re-establishment, lations of timber and non-timber benefits. Section three and Norway spruce is planted at a density of 4,500 trees per four present and discuss the results, respectively. hectare. We assume standard practice of a heavy thinning early in the stand’s life whereafter it develops into a light thinning once the stand attains a height makes it susceptible Materials and methods to windthrow (12–14 m) (Danish Forest Society 1995, cited by Nord-Larsen et al. 2003). Site description and the scenarios The ‘‘conversion scenario’’ is defined by principles of NNF (Larsen and Nielsen 2007) of beech, using a uniform The study focuses on converting the current RF of Norway shelterwood system with a regeneration thinning at stand spruce stand into NNF of beech based on two different sites age 90 years and liquidation of the last holdovers at age 123 Eur J Forest Res (2014) 133:611–622 613

110 for an infinite time horizon on the good soil. The TNPV ¼ NPVTimber þ NPVRecreation þ NPVWater corresponding shelterwood stand ages on the poor soil are extended to 100 and 130 (Forest Economics Table 2006). In the following subsections, the economic calculation At the time of establishment, beech is planted at a density of timber, recreation benefit, and groundwater benefits are of 5,500 trees per hectare and followed by continuous described. thinning, regeneration harvests, and initiation of natural regeneration. Economic calculation of timber benefit Methods The tree growth model outputs are based on a standard growth and yield table (Forest Economics Table 2006). The current study applied an economic cost-benefit ana- Economic data including regeneration costs and stumpage lysis (cf Hanley and Barbier 2009) by means of benefit price were found from Forest Economics Table (2006) and transfer to assess the monetary values of timber and non- adjusted for inflation to year 2012 (Statistics Denmark timber benefits. We used net present value (NPV) for an 2012). The regeneration costs for Norway spruce include infinite time horizon to evaluate the conversion (Davis soil preparation, plants, planting, and pesticides. The et al. 2001, p. 332). The NPV formula for timber, recrea- regeneration costs for conversion phase or initial stand tion, and water are described below. establishment of beech include fencing besides the other The NPV for an infinite time horizon of timber for RF is costs. The regeneration costs of beech differ between the calculated as the sum of the discounted future benefits and conversion phase and the post-conversion phase, as natural costs with a discount rate (r) for infinite rotations of length regeneration is possible in the latter due to the presence of R, using the following formula: seed trees, and also fencing cost is avoided. The stumpage R NPV ð1 þ rÞ price functions consider timber product categories/assort- NPVRF ¼ ð1 þ rÞR1 ment of beech and Norway spruce (‘‘Appendix B’’). The stumpage prices are net mill selling price after deducting The NPV for an infinite time of timber for NNF is harvesting, bunching, and transport costs. Stumpage price calculated as the sum of the present value of net revenue of beech ‘‘without red-heart formation’’ was used because (NR) of beech of the conversion phase and the present according to Nord-Larsen et al. (2003) shelterwood natural value of the expectation value of the final steady state as regeneration of beech in early years of liquidation has low follows: ! propensity to such damage. Turnover balances for both XR R NRt NPVðÞ 1 þ r R scenarios on the two soil conditions are given in NPVNNF ¼ t þ R ðÞ1 þ r ‘‘Appendix A’’. t¼0 ðÞ1 þ r ðÞ1 þ r 1

The NPV of an infinite time horizon of the marginal Estimation of recreation benefit increase in recreation benefit of the conversion into NNF is calculated as the sum of present value of recreation benefit The change in recreation benefit of converting RF of of conversion phase and present value of the infinite value Norway spruce stand into NNF of beech was estimated of recreation benefit of post-conversion using the above considering the influence of tree size on recreationalists’ formula for NPV of NNF. This is because the recreation- preferences. The value is based on Nielsen et al. (2007) alists’ WTP was estimated for an increase in WTP for who used the choice experiment method to elicit the changing from RF of Norway spruce stand to NNF of Danes’ WTP for RF of Norway spruce and NNF of beech beech stand (Nielsen et al. 2007). represented by different stand characteristics in a maturity Water yield data are available only as an annual average phase. The respondents’ preferences for changing from RF over a rotation (Duncker et al. 2011), and hence, the infi- with clear-felling, Norway spruce trees and no deadwood nite value of NPV of water benefit is calculated using trees into NNF with shelterwood, beech trees, and few annuity discounting factor and by discounting for water deadwood trees scenario were estimated at € 28, 103, and travel time (T) 20 years as follows: 15 per person per year, respectively. The NNF scenario Annual water benefit would thus consist of all three aspects i.e. € 146 per person NPV ¼ ðÞ1 þ r T per year or € 172 in 2012 prices. The authors argue that the r values might partly be rooted in the positive cognitive Total net present value (TNPV) for NNF and RF is value ascribed to nature-based forest management, and thus calculated as the sum of NPV of timber, NPV of recreation not solely relates to the recreational use value. We take up benefit, and NPV of water benefit. a conservative approach here and divide this number by

123 614 Eur J Forest Res (2014) 133:611–622

Table 2 Recreational benefits for the two locations once a mature Estimation of groundwater benefit NNF beech forest stand has been established Gribskov Klosterheden The demand for groundwater is high in northern Zealand and Jutland, where some portions of both case areas are classified Area (ha) 5,698 6,265 as a ‘‘particularly valuable’’ groundwater abstraction areas Visiting cars per year (2.5 persons per car; 179,200 154,250 (GEUS 2009). Hence, a monetary value of the gain in Jensen 2003) groundwater recharge of converting the RF of Norway spruce Visits per hectare 79 25 into NNF of beech was estimated considering the use value of WTP € ha-1 year-1 119 38 water for drinking purpose and the existence value of water in WTP is € 1.5 per visit Denmark. Duncker et al. (2011) found that NNF of beech increases water yield (run-off) by about 66 mm (660 m3 ha-1 year-1) above RF of Norway spruce. This run- three due to hypothetical bias (cf List and Gallet 2001), off difference is also assumed to represent Denmark’s forest which gives € 57. As the average Dane visits the forest 38 ecosystem (Raulund-Rasmussen 2012). Another study in times per year (Jensen and Koch 2004), the WTP per visit Denmark found that the evapotranspiration of mature conifers is € 1.5. Table 2 shows the resulting WTP for recreation forest exceeded evapotranspiration of mature forest per hectare for the two sites. The value of deadwood was by up to 35 mm year-1, in years 2001–2003 where the mean implicitly excluded in the first rotation or conversion phase annual precipitation was 886 mm (Boegh et al. 2009). by using a lower recreation score for conversion phase than Therefore, to be conservative, a water gain of post-conversion phase (see Table 3). 330 m3 ha-1 year-1 (half of 66 mm) was assumed in the The WTP per ha value was estimated at a maturity present study. Furthermore, the water conversion efficiency phase. Following the EFORWOOD project (Edwards from run-off state to groundwater recharge was discounted for et al. 2010), the forest was decomposed into an estab- 50 % water loss through percolation process, and so the water lishment, a young and a medium phase as seen in gain is 165 m3 ha-1 year-1. This efficiency rate is assumed to Table 3. The WTP was reduced for the respective phases reflect the average hydrological cycle of Denmark (Raulund- by dividing the recreational score by the score of the Rasmussen 2012). The water conversion efficiency was also maturity phase. WTP for the conversion phase was traced discounted for an average delay of water ‘‘travel time’’ of using recreation score for clear-felling of beech. WTP for 20 years (Loubier 2003). Klosterheden receives an annual post-conversion phase was traced using recreation score precipitation of 850 mm, whereas Gribskov receives 613 mm for NNF. Then, NPV of recreation for infinite time (DMI 2012). Nevertheless, the run-off difference between horizon was estimated using the NPV formula for NNF NNF and RF is approximately the same (Raulund-Rasmussen discussed in section ‘‘Methods.’’ 2012).

Table 3 Adjustment of recreation benefit using the recreation score of different phases of stand development of near-natural forestry of beech composed of conversion phase and post-conversion (Edwards et al. 2010,p.9) Phases Year Conversion phase Post-conversion Recreation Recreation score WTP Recreation Recreation score WTP score factor € ha-1 year-1 score factor € ha-1 year-1

(a) Gribskov: conversion phase (0–89 years) and post-conversion (90 years–infinity) Establishment 0–5 2 0.33 26 4 0.44 53 Young 5–15 3 0.50 40 5 0.56 66 Medium 15–50 5 0.83 66 7 0.78 92 Mature >50 6 1.00 79a 9 1.00 119 (b) Klosterheden: conversion phase (0–99 years) and post-conversion (100 years–infinity) Establishment 0–5 2 0.33 8 4 0.44 17 Young 5–15 3 0.50 12 5 0.56 21 Medium 15–50 5 0.83 21 7 0.78 29 Mature >50 6 1.00 25 9 1.00 38 For conversion phase, a recreation score of clear-felling of beech (Forest Management 4) and for post-conversion phase, a recreation score of near-natural forestry of beech (Forest Management 2) were assumed a Recreation benefit of post-conversion at maturity (119 € ha-1 year-1) times the ratio of recreation score of clear-felling of beech to recreation score of near-natural of beech at maturity (6/9)

123 Eur J Forest Res (2014) 133:611–622 615

Table 4 Groundwater benefit of converting rotational forestry (RF) account (€ per hectare) that allowed both aggregation of of Norway spruce into near-natural forestry (NNF) of beech consid- and comparison between the benefits. An exchange rate of -3 ering sum of drinking water price of 1.65 € m (DANVA’S 1 Euro (€)= 7.45 Danish Kroner in 2012 was assumed. The Benchmarking and Water Statistics 2010) and existence value of 1.12 € m-3 (Campbell et al. 2012) economic values were adjusted for inflation towards year 2012 using national consumer price index (Statistics Den- Items Value mark 2012). Future benefits and costs were discounted at a Run-off difference between NNF and 330 baseline discount rate of 3 %, and 1, 2, and 4 %. RF (m3 ha-1 year-1) (Duncker et al. 2011) Groundwater recharge (%) (Raulund-Rasmussen 2012)50 Groundwater recharge (m3 ha-1 year-1) 165 Results Groundwater benefit (€ ha-1 year-1) 456.5 Groundwater benefit assuming 20 years delay 252.7 The results are presented in three main sections. The first -1 -1 (€ ha year ) section presents developments of annual timber and non- timber benefits over time. The second section compares the NPV of timber benefit between NNF and RF in the two soil The annual groundwater benefit per ha was then calculated conditions under different discount rates. The third section by considering the use value and the existence value of water. presents the change in NPV of non-timber benefits for the As Hasler et al. (2005) argues, the value of clean groundwater conversion into NNF in the two site conditions and com- is to a large extent a value consisting of the knowledge of pares it with the change in NPV of timber. knowing that we have clean groundwater, as even without, it would just be cleaned. Thus, it is too simple to look at the use Development of annual recreation benefit value alone. We try to split the two in order to get a conser- vative estimate, as well as a more realistic one. The use value Figure 1 below shows the development of annual recrea- of groundwater benefit was calculated by multiplying the tion benefit of NNF over the conversion phase (0–90 years above groundwater recharge by the price of drinking water in in Gribskov and 0–100 in Klosterheden) and post-conver- Denmark. The average price of water as perceived by con- sion phase afterwards. The recreation benefit increases over sumer in Denmark is 7.02 € m-3. This price consists of 22 % time as it is a function of increasing recreational scores clean groundwater, 48 % waste water fee, and 30 % state from establishment phase to maturity phase. For equivalent taxes and VAT (DANVA’S Benchmarking and Water Sta- phases of stand development, the recreation benefit is tistics 2010). Hence, the price of groundwater in the current higher in the post-conversion phase than in the conversion study was estimated at 1.54 € m-3 (0.22 9 7.02), by phase, which reflects a higher recreation value of shelter- excluding the values of other utilities, in year 2009 and wood regime, and old and deadwood trees as compared adjusted to 1.65 € m-3 in 2012. The existence value was with initial uniform plantation, and no old and deadwood calculated by multiplying the above groundwater recharge by trees (Edwards et al. 2010). The recreation benefit in the WTP of Danish households for an increase in groundwater Gribskov is about 3 times higher than in Klosterheden at all recharge. A choice experiment survey undertaken in August– times, attributed to the number of visits. September2011byCampbelletal.(2012) estimated Danish household’s WTP for an increase in groundwater recharge Development of annual timber and non-timber benefits under forests by 40 million m3 at 53.82 € year-1. We divide this number by three due to hypothetical bias (cf List and The development of annual timber cash flow difference Gallet 2001), which gives € 17.94. This value was further between NNF and RF is compared with the development of converted into 1.12 m-3 by multiplying the household’s WTP annual non-timber benefits in Fig. 2a, b below. The annual to the ratio of total number of households in Denmark, 2.5 non-timber benefit development is the sum of annual rec- million (Statistics Denmark 2012) to the assumed increase in reation benefit and annuity water benefit of 457 € ha-1 for groundwater recharge, 40 million m3. use value and existence value, considered after 20 years of The use value alone was considered as a lower-bound water travel time. Figure 2a shows the developments for value in the CBA, and the sum of use value and existence Gribskov over the first two rotations (0–179 years), and value was considered as a higher-bound value (Table 4). Fig. 2b shows the developments for Klosterheden over the first two rotations (0–199 years). Positive cash flows of Relevant assumptions of CBA timber indicate the gain from the conversion, and negative cash flows indicate the opportunity cost of the conversion, The economic calculations of timber, recreation, and particularly from final harvesting of RF. When comparing groundwater benefits were made on a common unit of timber cash flow between the two figures, the conversion to 123 616 Eur J Forest Res (2014) 133:611–622

Fig. 1 Development of annual Development of annual recreation benefit of near natural-forestry in recreation benefit of the Gribskov and Klosterheden conversion into near-natural 140 forestry of beech in Gribskov Klosterheden Gribskov and Klosterheden, Denmark. 120 The vertical break lines at year 100 90–99 and 100–109 are the conversion ages in Gribskov 80 and Klosterheden, respectively 60

Euro/year/ha 40

20

0

Year

Fig. 2 Developments of annual (a) Development of annual timber net cash flow and non-timber benefits of the timber net cash flow and non- conversion into near-natural forestry in Gribskov timber benefits of the 2,000 conversion to near-natural Timber Recreation + water forestry of beech in Gribskov (a) and Klosterheden (b). 1,500 Annual timber cash flow is the difference between near-natural forestry of beech and rotational 1,000 forestry of Norway spruce. Water benefit of use value and existence value added to 500 recreation benefit after 20 years of water travel time Euro/year/ha 0

-500

-1,000 Year

(b)Development of annual timber net cash flow and non-timber benefits of the conversion into near-natural forestry in Gribskov 800 Timber Recreation + water 600

400

200

0 Euro/year/ha -200

-400

-600 Year

123 Eur J Forest Res (2014) 133:611–622 617

Table 5 Net present value (€ ha-1) of timber of NNF of beech and of NNF is positive at 1 % discount rate but is still by far RF of Norway spruce in Gribskov and Klosterheden under different lower than the NPV of RF. discount rates Discount Gribskov (good soil) Klosterheden (poor soil) rate (%) Present values of non-timber benefits NNF RF NNF– NNF RF NNF– RF RF Table 6 shows the change in present values of non-timber 1 27,962 12,933 15,029 2,528 7,922 -5,394 benefits for the conversion from RF into NNF at the two 2 2,987 2,801 186 -5,866 562 -6,428 site conditions at a baseline discount rate, 3 %, and at a low 3 -3,738 -362 -3,375 -7,872 -1,666 -6,206 discount rate, 1 %. The difference in NPV of timber 4 -6,279 -1,794 -4,484 -8,504 -2,631 -5,874 between NNF and RF is also included in the table to compare it with that of non-timber benefits. In Gribskov, the conversion into NNF would gain an NPV of 6,832 € ha-1 from the recreation and the use values Table 6 Differences in net present value (€ ha-1) of timber, recre- of water benefit at a 3 % discount rate. This value is large ation, and water benefits between near-natural forestry (NNF) of enough to compensate for the loss of timber, which equals beech and rotational forestry (RF) of Norway spruce in Gribskov and 3,375 € ha-1. However, in Klosterheden, the conversion Klosterheden gains an NPV of 5,581 € ha-1 from recreation and water Benefits Net present value (NPV) for perpetuity (NNF–RF) benefit which is 1,251 € ha-1 less than that in Gribskov (i.e. Gribskov Klosterheden attributed to difference in recreation benefit). This is worth 625 € ha-1 less than the loss of timber. However, when 1% 3% 1% 3% considering the sum of use value and existence value of Timber 15,029 -3,375 -5,394 -6,206 water, the worth would be large enough to over-compensate Recreation 7,366 1,822 2,310 571 the loss of timber. Also, the value of non-timber benefits of Water 22,249 5,010 22,249 5,010 the conversion would be by far larger than the cost of tim- (37,410) (8,425) (37,410) (8,425) ber, giving a net worth of 19,165 € ha-1 at 1 % discount Recreation and 29,614 6,832 24,558 5,581 rate. The corresponding value in Gribskov is € 44,643. water Total 44,643 3,457 19,165 -625 The values of water in the parenthesis indicate the higher-bound Concluding discussion values when considering the sum of use values and existence value of water This study assessed the welfare economic values of timber and non-timber benefits of the conversion of RF of Norway NNF gains a significantly larger annual net revenues in spruce into NNF of beech based on two site conditions in Gribskov, during the period of 90–110 years where Denmark. We find that the economic value of use values of regeneration harvests initiated, than in Klosterheden in the recreation and water benefits outweighs the loss of timber corresponding periods. The difference in average annual on good soil conditions but not on poor soil conditions. benefit between timber and non-timber benefits is higher in This means that from a welfare economic point of view, it Klosterheden than in Gribskov. The average annual gain will be beneficial to convert forests on the better soils, but from timber and non-timber in Gribskov, being 234 and obviously it does not mean that the full surplus should be 488 € ha-1, respectively (Fig. 2a), whereas, in Klosterhe- spend on e.g. transfer to landowners for providing these den being minus 14 and 437 € ha-1, respectively (Fig. 2b). services. As always, efficiency should be sought. The result of a cost-benefit analysis like the one provided here hinges Net present value of timber on the underlying assumptions. In particular, the stated preference methods have been criticized (see e.g. Price Table 5 presents profitability or NPV of NNF and RF in 1996). Nevertheless, they are the only way to estimate non- Gribskov and Klosterheden under different discount rates. use values, and ignoring these would clearly make flaws to In Gribskov, NNF would be more profitable than RF at 1 the CBA. Therefore, we have been very conservative in the and 2 % discount rates. However, NNF of beech would be estimation of timber, and recreation and water benefits are more costly than RF of Norway spruce at 3 and 4 % dis- based on conservative assumptions. In the following, we count rates. In Klosterheden, NNF would be by far more discuss the arguments for where we have been most costly than RF at 2 % and higher discount rates. The NPV conservative.

123 618 Eur J Forest Res (2014) 133:611–622

Estimates of timber and non-timber benefits of near- timber value (Fig. 2; Table 6). This evidence reinforces the natural forestry argument of Wobst (2006). He argued that NNF sounds a promising strategy to overcome the economic crisis of First of all, the estimates of NPV of timber for converting forestry, caused by lower timber prices and higher costs, by RF into NNF is conservative since the negative impact of claiming a fair payment for environmental and recreational storm risk on the RF is ignored. The vulnerability of RF of functions of NNF. As shown in Fig. 2, the conversion into Norway spruce to storm risk is a serious problem, partic- NNF provides a stable flow of non-timber benefits to ularly in the Atlantic region of Europe (Vedel et al. 2010). compensate for part of the regeneration costs and supports For instance, in Denmark, a mean annual storm probability the positive net revenues. The current study further shows of 8 % is assumed in RF of Norway spruce, which causes that the contribution of non-benefits is significantly associated wind throw problem especially for larger dependent on site attributes. While on a good soil condi- diameter classes (Østergaard 1988, cited by Schou et al. tions and high visitor frequency site, the sum of NPV of use 2012). values of recreation and water benefits is capable of com- Danes’ € WTP of 1.5 per visit (Nielsen et al. 2007) for pensating for a modest loss of timber benefits, in a poor soil changing from RF of Norway spruce into NNF of beech is and low visitors frequency site, the NPV of use values of assumed to be conservative. Because the estimate ignores recreation and water benefits is not large enough to over- some positive recreational values of NNF such as off-site compensate a high loss on unproductive soil (Table 6). The visual experiences of landscape (Larsen and Nielsen 2007) loss of timber in poor soil is over-compensated by and non-use values (Zandersen et al. 2007; Campbell et al. including the existence values of water in addition to the 2012), and negative recreational values of RF of Norway use values of recreation and water benefit. Some studies spruce such as fresh windthrown trees (Lindhagen and found that the recreational value of NNF is substantial and Ho¨rnsten 2000) and application of chemicals (Jensen and capable of compensating for the loss of timber. For Koch 2004). Prokofieva et al. (2011) reviewed that in instance, Bostedt and Mattsson (2006) evidenced that Denmark forest, WTP per visit on a low level is € 1.1, a transformation from the RF-dominant forest landscape into medium level is € 4.8, and a high value is € 10.5 in 2005. a more recreation-oriented NNF in the county of Vaster- More importantly, the estimated recreation benefit is very botten (Sweden) would yield a three times larger NPV realistic since the WTP estimate was adjusted for change in from recreation benefit than the loss of timber discounted recreational values over different phases of stand devel- for 100 years of rotation at a 3 % discount rate. The study opment (Edwards et al. 2010). Estimation of water benefit applied an average Swedish recreationalist annual WTP of is conservative because run-off difference between NNF of € 134 per person, which was not corrected for a hypo- beech and RF of Norway spruce (330 m3 ha-1 year-1), thetical bias. However, in the current study, the recreation which is half of the estimate of Duncker et al. (2011), was benefit alone is not large enough to over-compensate for considered due to high uncertainties with change in climate the loss of timber even at a 1 % discount rate, mainly factors in the future. Change in climate factors in the because the WTP is corrected for hypothetical bias future, especially prediction of warmer and drier summers, (Table 6). will likely put pressure on water supplies in Europe (Pro- In a meta-analysis, List and Gallet (2001) found that kofieva et al. 2011). Additionally, groundwater recharge WTP from hypothetical stated values is overstated by three was considered as 50 % of the run-off and discounted for a as compared with the actual values. This was in the current 20-year water travel time (Loubier 2003). However, the study used to correct all stated preference study values. estimation of present value of water benefit is likely to be This may be seen too as a conservative approach, but is slightly overestimated as compared with the present values made here to avoid criticism of using stated preference of recreation and timber benefits since it was calculated by results in a CBA. applying an annuity factor (i.e. dividing water benefit by It should be noted that incorporating the welfare eco- discount rate). This is due to lack of available run-off nomic values of non-use benefits, such as biodiversity database over a rotation, where only the average run-off is conservation in the current cost-benefit analysis, would available, in Duncker et al. (2011). Nevertheless, the run- magnify the value of NNF policy. Duncker et al. (2011) off is stable over a rotation, meaning the uncertainty on show that the biodiversity score of NNF of beech is 129, present value would not be as high as in recreation. whereas that of RF of Norway spruce is one. The biodi- Based on such conservative estimates, the study shows versity score was based on important attributes such as the that the non-timber benefit of NNF is superior to RF, which abundance of deadwood, number of big trees, and number illuminates the importance of providing adequate incen- of tree species. From the demand side, Jacobsen et al. tives for provision of non-marketed goods such as recrea- (2012) provide evidence that Danish households are willing tion and water benefit of NNF to compensate for the loss of to pay up to 80 € year-1 for increasing the population of 123 Eur J Forest Res (2014) 133:611–622 619 endangered species in Denmark’s forests. NNF may be development type’’ (FDT), such as open lands, broad- better than RF at securing that. leaved-dominant FDT, and conifers-dominant FDT, would A challenge for an economic assessment of timber and increase the attraction value at the landscape level (Danish non-timber benefits in a multi-purpose NNF is the trade- Ministry of the Environment, Forest and Nature Agency offs between the benefits. In this regard, an important 2005; Price and Price 2008). Another important limitation methodological contribution of the present study is to of stand-level recreation benefit estimate is using per address trade-offs between timber production and recrea- hectare metric. Calculating number of visitors per hectare tion provision. The average recreationalists in Denmark in assumes uniform distribution of visitors to each hectare of particular and in Europe in general have shown low pref- the forest. However, practically each FDT in the forest erences for heavy interventions during the establishment receives different recreationalists depending on their attri- and young phases of the forest stand (Edwards et al. 2010). butes. In the present case, for example, NNF of beech By employing low recreation score for such phases of stand become bare and unattractive during winter season, and development (Edwards et al. 2010), it was possible to may receive very low visits and so loss of recreation value. translate the effect of timber production activities into a The economic valuation of timber, recreation, and water reduction in recreation benefit (Fig. 1). benefits might be higher than in many other countries, given relatively higher socio-economic conditions of Denmark Limitations of the study than the average of European countries (OECD 2012).

The stand-level estimates of timber and non-timber benefits Acknowledgments Jette Bredahl Jacobsen and Thomas Lundhede of the current study, estimated per ha, can be used be to would like to thank the Danish National Research Foundation for supporting the research at the Center for Macroecology, Evolution extrapolate a forest value in the study areas or elsewhere. and Climate. However, when extrapolating the results, the user needs to consider specification of the baseline and conversion sce- narios of forest management, limitations of the stand-level Appendix analysis, and the variations of socio-economic and site characteristics. Appendix A: Net revenue of timber in Gribskov The current study is based on a generic European forest and Klosterheden management regime classification(Raulund-Rasmussen et al. 2011) where the baseline scenario, RF of Norway See Tables 7, 8, 9, 10, 11 and 12. spruce, represents FM41 (forest management) and the conversion scenario, NNF of beech, represents FM2. NNF Appendix B: Stumpage price curves of beech is based on uniform shelterwood system. Some studies have shown that the recreational value of NNF of See Fig. 3. beech using uniform shelterwood system is lower than that of selective logging system (Nielsen et al. 2007). Timber benefit is also found to be lower in shelterwood than in Table 7 Calculation of net revenue of beech for conversion phase in Gribskov, site class 1 selective logging system (Nord-Larsen et al. 2003; Price and Price 2008). Water run-off is relatively higher in Age Regeneration Harvest DBH Stumpage Net cost (€) (m3) (cm) price revenue shelterwood system than in selective logging due to a (€ m-3) (€ ha-1) higher water evaporation loss in a continuous vegetation cover in selective logging (Dalsgaard 2008; Katzensteiner 2 6,713 -6,713 et al. 2011). Therefore, such differences in timber and non- 3–9 2,183 2,183 timber benefits between shelterwood system and the 10–19 1,097 6 5.5 14.7 -1,009 alterative, selective logging system of NNF, need to be 20–29 29 9.1 23.3 676 considered when using the alternative system. 30–39 72 13.0 30.3 2,182 Making an economic estimate of recreation benefit at 40–49 73 18.1 29.6 2,162 stand level in the present study would conceal an actual 50–59 73 23.7 29.1 2,122 value generated as a result of spatial interaction or 60–69 74 28.1 29.7 2,197 arrangement of different stands in the forest landscape. 70–79 72 33.0 34.1 2,453 This is because the mosaic of different ‘‘forest 80–89 68 36.7 39.1 2,656 The regeneration cost is for planting of beech in the initial stand 1 The number in the forest management regime indicates the intensity establishment phase. Based on data from Forest Economics Table of silvicultural intervention in an increasing order. (2006) 123 620 Eur J Forest Res (2014) 133:611–622

Table 8 Calculation of net revenue of near-natural forestry of beech for post-conversion phase in Gribskov, site class 1 Age Thinning Regeneration harvest Regeneration Harvest DBH Stumpage price Net revenue Harvest DBH Stumpage price Net revenue cost (€) (m3) (cm) (€ m-3) (€ ha-1) (m3) (cm) (€ m-3) (€ ha-1)

0/90 837 -837 86 43.9 53.1 4,556 0–9/90–99 123 46.6 60.3 7,408 10–19/100–109 600 6 5.5 14.7 -512 105 52.0 62.5 6,573 110 0 276 55.0 72.5 19,994 20–29 1,010 29 9.1 23.3 -334 30–39 72 13.0 30.3 2,182 40–49 73 18.1 29.6 2,162 50–59 73 23.7 29.1 2,122 60–69 74 28.1 29.7 2,197 70–79 72 33.0 34.1 2,453 80–89 68 36.7 39.1 2,656 The regeneration cost is for the natural regeneration of beech. Based on data from Forest Economics Table (2006)

Table 9 Calculation of net Age Regeneration cost Harvest DBH Stumpage Net revenue revenue of rotational forestry of (€) (m3) (cm) price(€ m-3) (€ ha-1) Norway spruce in Gribskov, site class 2 4 3,207 -3,207 5–9 612 -612 10-19 105 14 7.2 1.7 -81 20–29 82 10.2 3.3 271 30–39 119 15.0 7.0 836 40–49 132 20.7 14.3 1,898 50–59 133 26.7 23.4 3,111 Based on data from Forest 60 377 34.1 25.3 9,543 Economics Table (2006)

Table 10 Calculation of net Age Regeneration cost Harvest DBH Stumpage price Net revenue revenue of beech for conversion (€) (m3) (cm) (€ m-3) (€ ha-1) phase in Klosterheden, site class 4 2 6,713 -6,713 3–9 2,183 -2,183 10–19 1,097 -1,097 20–29 6 3.2 8.4 50 30–39 17 6.9 18.4 313 40–49 24 9.9 25.0 599 50–59 30 13.0 30.3 894 60–69 35 15.7 29.9 1,048 The regeneration cost is for planting of beech in the initial 70–79 35 18.7 29.5 1,034 stand establishment phase. 80–89 36 21.8 29.1 1,049 Based on data from Forest 90–99 43 25.2 29.1 1,237 Economics Table (2006)

123 Eur J Forest Res (2014) 133:611–622 621

Table 11 Calculation of net revenue of near-natural forestry of beech for post-conversion phase in Klosterheden, site class 4 Age Thinning Regeneration harvest Regeneration Harvest DBH Stumpage price Net revenue Harvest DBH Stumpage price Net revenue cost (€) (m3) (cm) (€ m-3) (€ ha-1) (m3) (cm) (€ m-3) (€ ha-1)

0/100 837 -837 48 28.9 30.3 1,449 0–9/100–109 62 32.0 33.3 2,052 10–19/110–119 600 -600 66 35.5 37.8 2,497 20–29/120–129 1,010 6 3.2 8.4 -960 62 40.3 46.8 2,887 130 142 44.8 54.1 7,673 30–39 17 6.9 18.4 313 40–49 24 9.9 25.0 599 50–59 30 13.0 30.3 894 60–69 35 15.7 29.9 1,048 70–79 35 18.7 29.5 1,034 80–89 36 21.8 29.1 1,049 90–99 43 25.2 29.1 1,237 The regeneration cost is for the natural regeneration of beech. Based on data from Forest Economics Table (2006)

Table 12 Calculation of net Age Regeneration cost Harvest DBH Stumpage price Net revenue revenue of rotational forestry of (€) (m3) (cm) (€ m-3) (€ ha-1) Norway spruce in Klosterheden, site class 3 4 3,207 -3,207 5–9 612 -612 10–19 105 -105 20–29 40 7.7 1.9 75 30–39 80 11.4 4.1 326 40–49 90 15.7 7.7 692 50–59 131 20.7 14.3 1,868 60–69 78 25.8 22.1 1,724 Based on data from Forest 70 378 29.1 24.9 9,419 Economics Table (2006)

Fig. 3 Stumpage price of beech and Norway spruce: Based on data from Forest Economics Table (2006)

123 622 Eur J Forest Res (2014) 133:611–622

References List JA, Gallet CA (2001) What experimental protocol influence disparities between actual and hypothetical stated values? Evi- Boegh E, Poulsen RN, Butts M, Abrahamsen P, Dellwik E, Hansen S, dence from a meta analysis. Environ Resour Econ 20:241–254 Hasager CB, Ibrom A, Loerup J-K, Pilegaard K, Soegaard H Loubier S (2003) Economic assessment of groundwater protection: a (2009) Remote sensing based evapotranspiration and runoff sensitivity analysis of costs-benefits results illustrated by a small modeling of agricultural, forest and urban flux sites in Denmark: aquifer protection in North Jutland region, Denmark. Case study from field to macro-scale. J Hydrol 377:300–316 report no.3 - BRGM/RC: 52326-FR. BRGM, Orle´ans Boman M, Jacobsen JB, Strange N, Norman J, Mattsson L (2010) Møller CM (1933) Boniteringstabeller og bonitetsvise tilvækstover- Forest amenity values and the rotation age decision: a Nordic sigter for Bøg, Eg Og Rødgran I Danmark [Site indexcurves for perspective. Ecol Bull 53:7–20 yieldtables for beech, oak and Norwayspruce in Denmark. Dansk Bostedt G, Mattsson L (2006) A note on benefits of adjusting forestry Skovbrugs Tidsskrift 18:537–623 (in Danish) to meet recreational demands. J For Econ 12:75–81 Nielsen AB, Olsen SB, Lundhede TH (2007) An economic valuation Campbell D, Suzanne E, Vedel SE, Jacobsen JB, Thorsen BJ (2012) of the recreational benefits associated with nature-based forest Heterogeneity in the demand for recreational access–distribu- management practices. Landsc Urb Plan 80:63–71 tional aspects. Paper accepted for presentation at the EAERE Nord-Larsen T, Bechsgaard A, Holm M, Holten-Andersen P (2003) conference in Prague, June 2012, p 19 Economic analysis of near-natural beech stand management in Dalsgaard L (2008) Transpiration and water budgets of European Northern Germany. For Ecol Manag 184:149–165 beech (Fagus sylvatica L.) dominated stands in relation to Price C (1996) Contingent valuation and retrograde information bias. canopy structure. Forest and landscape Research No. 39-2008. In: Park A, Stewart Roper C (eds) The living forest. Proceedings Forest and landscape Denmark. Frederiksberg of the International Symposium on the non-market Benefits of Davis LS, Johnson KN, Bettinger P, Howard TE (2001) Forest Forestry, Edinburgh, June 1996, TSO, London, pp 37–44 management: to sustain ecological, economic and social values, Price C, Price M (2008) Cost-benefit analysis of continuous cover 4th edn. Waweland Press Inc., Illinois, p 804 forestry. In: Bergsend E, Delbeck G, Hoen HF (eds) Proceedings Duncker PH, Raulund-Rasmussen K, Gundersen P, Jong J, Katzensteiner of the biennial meeting of the scandinavian society of forest K, Ravn HP, Smith M, Otto Eckmu¨llner O (2011) Synergies and economics Lom, Norway, 6th–9th April, 2008. Scand For Econ trade-offs between production, land expectation value and ecolog- 42: 36–65 ical services like water, carbon sequestration, biodiversity, and soil Prokofieva I, Lucas B, Thorsen BJ, Carlsen K (2011) Monetary values fertility in relation to forest management. Ecol Soc ES-2011-4081 of environmental and social externalities for the purpose of cost- Edwards D, Jay M, Jensen FS, Lucas B, Marzano M, Montagne C, benefit analysis in the EFORWOOD project. EFORWOOD Peace A, Weiss G (2010) Assessment of the recreational value of Deliverable D1.5.6.CTFC, Spain European forest management alternatives. EFORWOOD Report Raulund-Rasmussen K (2012) Personal communication. Forest & D2.3.6, Forest Research, UK Landscape, University of Copenhagen, Rolighedsvej 231958 Hanley N, Barbier EB (2009) Pricing nature: cost-benefit analysis and Frederiksberg C, on 12 April, 2012 environmental policy-making. Edward Elgar Publishing Limited, Raulund-Rasmussen K, Hansen K, Katzensteiner K, Loustau D, de Cheltenham Jong J, Gundersen P, Humphrey JW, Ravn HP, Klimo E (2011) Hasler B, Lundhede T, Martinsen T, Neye S, Schou JS (2005) Introduction. In: Raulund-Rasmussen K, Hansen K (eds) Syn- Valuation of groundwater protection versus water treatment in thesis report on impact of forest management on environmental Denmark by choice experiments and contingent valuation. NERI services. EFORWOOD Deliverable D2.2.2. Forest and Landsc, Technical Report no. 543 University of Copenhagen, Denmark (UC), pp 5–10 Jacobsen JB, Mo¨hring B, Wippermann C (2004) Business economics Rune F (2009) Gribskov. Forlaget ‘‘Esrum Sø’’, Fredensborg, Denmark of conversion and transformation—a case study of Norway Schou E, Jacobsen JB, Kristensen KL (2012) An economic evaluation of spruce in Northern Germany. In: Spiecker H, Hansen J, Klimo E, strategies of transforming even-aged into near-natural forestry in a Skovsgaard JP, Sterba H, von Teuffel K (eds) Norway spruce conifers-dominated forest in Denmark. For Policy Econ 20:89–98 conversion-options and consequences. European Forest Institute Spiecker H, Hansen J, Klimo E, Skovsgaard JP, Sterba H, von Teuffel Research Report 18. Brill Leiden, Boston, pp 225–252 K (2004) Summarizing discussion. In: Spiecker H, Hansen J, Jacobsen JB, Lundhede TH, Thorsen BJ (2012) Valuation of wildlife Klimo E, Skovsgaard JP, Sterba H, von Teuffel K (eds) Norway population above survival. Biodivers Conserv 21:543–563 spruce conversion-options and consequences. European Forest Jensen FS (2003) Friluftslivi 592 skove og andre naturomra˚der Institute Research Report 18. Brill Leiden, Boston, pp 252–260 [Outdoor recreation in 592 forests and other natural areas]. Tarp P, Helles F, Holten-Andersen P, Larsen JB, Strange N (2000) Skovbrugsserien nr.32, Skov and Landskab, Hørsholm (in Danish) Modelling near-natural silvicultural regimes for beech—an Jensen FS, Koch NE (2004) Twenty-five years of forest recreation economic sensitivity analysis. For Ecol Manag 130:187–198 research in Denmark and its influence on forest policy. Scand J Vedel SE, Jacobsen JB, Boon TE (2010) Atlantic urbanized region. For Res 19(4):93–102 In: NEWFOREX (New Ways to Value and Market Forest Katzensteiner K, Klimo E, Szukics U (2011) The impact of forest Externalities). D2.1 A Report describing the role of key management on the water cycle of forest stands and watersheds. externalities across case studies. METLA (Finnish Forest In: Raulund-Rasmussen K, Hansen K (eds) Synthesis report on Research Institute), pp 47–74 impact of forest management on environmental services. Wobst H (2006) Combination of economics and ecological aspects by EFORWOOD Deliverable D 2.2.2. Forest and landscape, close to nature forestry—a contribution to the economic crisis of University of Copenhagen, Denmark (UC), pp 55–72 forestry. In: Diaci J (ed) Nature-based forestry in Central Europe: Larsen JB, Nielsen AB (2007) Nature-based—where are we going? An alternative to industrial forestry and strict preservation. Studia Elaborating forest development type in and with practice. For Forestalia Slovenica 126. Department of Forestry and Renewable Ecol Manag 238:107–117 Forest Resources-Biotechnical Faculty, Ljubljana, pp 79–90 Lindhagen A, Ho¨rnsten L (2000) Forest recreation in 1977 and 1997 Zandersen M, Termansen M, Jensen FS (2007) Evaluating approaches in Sweden: changes in public preferences and behaviour. to predict recreation values of new forest sites. J For Econ Forestry 73:143–151 13:103–128

123